Learning new words

Abstract

Systems and methods are disclosed for a server learning new words generated by user client devices in a crowdsourced manner while maintaining local differential privacy of client devices. A client device can determine that a word typed on the client device is a new word that is not contained in a dictionary or asset catalog on the client device. New words can be grouped in classifications such as entertainment, health, finance, etc. A differential privacy system on the client device can comprise a privacy budget for each classification of new words. If there is privacy budget available for the classification, then one or more new terms in a classification can be sent to new term learning server, and the privacy budget for the classification reduced. The privacy budget can be periodically replenished.

Description

RELATED APPLICATIONS[0001]This application claims priority under 35 U.S.C. §119(e) of U.S. Patent Application No. 62 / 348,988, filed Jun. 12, 2016, and entitled, “LEARNING NEW WORDS,” and U.S. Patent Application No. 62 / 371,657, filed Aug. 5, 2016, entitled “LEARNING NEW WORDS,” both of which are incorporated herein by reference to the extent that they are consistent with this disclosure.[0002]This application is related to U.S. patent application Ser. No. 15 / 275,357, filed Sep. 24, 2016, and entitled “LEARNING NEW WORDS,” which is incorporated herein by reference to the extent that is it consistent with this disclosure.TECHNICAL FIELD[0003]This disclosure relates to the field of a server learning new words generated on a client device.BACKGROUND[0004]A user of a client device relies on one or more dictionaries of words for spell checking, suggesting words during typing, and other uses of known words. Such client dictionaries are difficult to keep updated with new words that may becom...

AI Technical Summary

Benefits of technology

The patent describes a system that allows a client device to determine if a new word is a new word that is not in a dictionary or asset catalog on the client device. These new words can be associated with different classifications such as entertainment, health, finance, etc. The system uses a differential privacy system to protect the privacy of the new words. The system can store more words in a transmission buffer than there is privacy budget available to transmit the words. The system can also use n-grams to introduce randomness in the data prior to sharing it with a server that will learn the new words. The server can test the differentially private data to determine if the amount of randomization is sufficient to maintain the privacy of the client data. The technical effects of this system include improved privacy and security for new words and better data management.

Problems solved by technology

A server can learn from the aggregation of the crowdsourced data of all clients, but the server cannot learn the data provided by any particular client.

Images